Fuel cell ship power system fault reconstruction method

Abstract

The invention discloses a fuel cell ship power system fault reconstruction method, which comprises the steps of building a fuel cell ship annular power system simulation model, designing an improved particle swarm algorithm, initializing a candidate solution by using a logic mapping chaos model, updating the candidate solution through a quantum particle swarm algorithm, and introducing a diversity migration strategy, particles in different ranges in the population are captured, migration individuals are selected based on the fitness and population position information, the individual with the minimum average Hamming distance in the population indicates the optimization direction of the iteration population, and the improved particle swarm algorithm is integrated into the fuel cell ship annular power system simulation model. The maximum recovery load, the minimum switching operation frequency and the maximum power supply load rate of each area serve as objective functions, optimization solution is carried out on the basis that weighted summation is converted into a single objective function, a power supply network under the fault condition is reconstructed through an improved particle swarm algorithm, a reconstruction scheme is obtained, and power system reconstruction after the fault is completed.

Description

TECHNICAL FIELD

[0001] The application relates to a fuel cell ship power system, in particular to a fuel cell ship looped network type power system fault reconstruction method based on an improved particle swarm algorithm. BACKGROUND

[0002] With the access of new energy such as solar energy, wind energy and fuel cell and the development of ship intelligence, more and more intelligent technologies and power electronic devices are integrated into the existing ship infrastructure, which continuously increases the complexity and heterogeneity of the ship power system. When the ship power system fails, it may reduce the reliability of the system, or even cause serious damage to the equipment and even cause power supply path blockage, which cannot guarantee the smooth sailing and personnel safety. Therefore, in order to meet the needs of modern technology, the fuel cell ship looped power system needs an intelligent reconstruction technology that can detect faults in time and reconstruct and recover the power system to normal operation in time when the ship fails SUMMARY In view of the slow convergence speed and easy falling into local optimal solution of the existing ship power system fault reconstruction algorithm, a fuel cell ship power system fault reconstruction method based on an improved particle swarm algorithm is provided, which specifically comprises the following steps: A fuel cell ship looped power system simulation model is built, the ship loads are classified by importance level, and the key loads are configured with standby power supply paths to control the continuity of power supply. The fuel cell ship looped power system simulation model simulates branch faults and power supply faults of the power system, and tests various fault conditions. An improved particle swarm algorithm is designed: a candidate solution is initialized using a logical mapping chaos model, the candidate solution is updated through a quantum particle swarm algorithm, a diversity migration strategy is introduced to capture particles of different ranges in the population, and a migration individual is selected based on fitness and population position information. The individual with the smallest average Hamming distance in the population indicates the optimization direction of the iteration population, and the particles deviating from the population center range are replaced by comparing the fitness value and the average Hamming distance between the particles. The improved particle swarm algorithm is integrated into the fuel cell ship looped power system simulation model, wherein each particle represents a solution to the fault reconstruction problem, and the position of each particle in space corresponds to a specific set of switch states. The first objective function, the second objective function and the third objective function are established respectively to maximize the recovered load, minimize the number of switch operations and maximize the power supply rate of each region, and the weighted sum of the three objective functions is converted into a final objective function and optimized. The improved particle swarm algorithm is used to reconstruct the power supply network under fault conditions to obtain a reconstruction scheme and complete the power system reconstruction after the fault. Further, when designing the improved particle swarm algorithm, The initial candidate solution is generated based on the logic mapping chaos model, and the uniformity of the candidate solution is improved:

[0003] In the formula is the candidate solution set, is the total number of candidate solutions, is the load sequence number.

[0004] Further, the position of the particle in the improved particle swarm algorithm is described using the Monte Carlo random equation, and the position equation of the particle in the potential well is obtained based on simulation and sampling:

[0005] Wherein is a random number uniformly distributed on [0, 1], is the characteristic length of the potential well, when the particle is in the multi-dimensional target search space, the related variable equation is updated, and then the position equation of the transformed particle is established as:

[0006] is the current best position of the th particle, is the contraction-expansion coefficient, is the average best position.

[0007] Further, when the diversity migration strategy is introduced: calculate the average Hamming distance in the population, select the particle with the smallest average Hamming distance as the migration object, and the average Hamming distance calculation formula of the particle is:

[0008] represents the total number of particles, and represent two particles.

[0009] According to the group representation, select the particle with the smallest average Hamming value and add it to the migration target list. When there are no two or more particles with the same average Hamming in the population, select the particle with the smallest average Hamming as the migration candidate particle; when there are two or more particles with the same average Hamming, compare the fitness values between the particle groups, and select the worst particle as the migration object.

[0010] Further, while increasing the diversity of the population, the worst particle is removed, and the particle with the largest fitness value is used to replace the determined migration object.

[0011] Furthermore, based on load priority, critical loads should be kept active, and third-level loads should be kept active as much as possible. Therefore, the first objective function is established as follows:

[0012] In the formula These are the active power values ​​for primary, secondary, and tertiary loads, respectively. These represent the number of load levels at different rates. This represents the power supply status of the load; 0 indicates the load is unloaded, and 1 indicates the load is loaded. These are the weights assigned to different load levels, with critical loads required to have significantly higher weights than other loads. The weighting expression is as follows:

[0013] Based on load priority, if the number of switching operations is kept to a minimum, then the second objective function is established as follows:

[0014] In the formula and Or 1, and This indicates the switch operation status and the total number of critical loads. This indicates that the number of three loads is determined by... express; To balance the load factor of the power supply in each area, a third objective function is established based on this condition:

[0015] The number of power sources in the formula is... express, express The actual power of the No. 1 distribution board express Rated power of the No. 1 distribution board; Construct the final objective function based on the first, second, and third objective functions:

[0016] In the formula These are the weight coefficients for each sub-objective, and the requirements are... , This represents the maximum value of the objective function.

[0017] According to the technical scheme, the fuel cell ship electric power system fault reconstruction method can perform fault reconstruction performance test based on fuel cell ship annular electric power system simulation, comprehensively consider load priority, switch operation number and each area power supply load rate balance to establish an objective function of electric power system reconstruction, introduce a diversity migration strategy on the basis of a quantum particle swarm algorithm, design an improved particle swarm algorithm for optimization process of fuel cell ship annular electric power system fault reconstruction, and has the advantages of fast convergence speed and high precision compared with a traditional intelligent optimization algorithm. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0019] Figure 1 The fuel cell ship annular micro-grid structure diagram of the present application; Figure 2 The improved particle swarm algorithm flowchart in the method of the present application; Figure 3 The fuel cell ship annular micro-grid system topology diagram of the present application; Figure 4 The fault reconstruction test result of the improved particle swarm algorithm of the present application; Figure 5 The fitness iteration process of the improved particle swarm algorithm of the present application; Figure 6 The distribution board load rate iteration process in the present application. DETAILED DESCRIPTION

[0020] In order to make the technical solutions and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application: The application discloses a fuel cell ship power system fault reconstruction method, which is based on a fuel cell ship annular power system to establish a fuel cell ship annular power system simulation model, wherein ship loads are classified according to their importance, which not only helps to reasonably schedule different loads, but also ensures that key loads can obtain power support preferentially when the system fails. In order to ensure the continuity of power supply of the system when the system fails, a backup power supply path is configured for important loads, so that the redundancy backup of the system is realized, and the reliability and fault tolerance of the power system are enhanced. On the basis of the system model, an improved particle swarm algorithm is used for fault reconstruction optimization. Particles are used as solutions to the fault reconstruction problem, and through updating iteration, an optimal reconstruction scheme is found, and the reconstruction of the power system after the failure is completed. Figure 1 As shown in the figure, the improved algorithm improves its adaptability to the problem of ship power system fault reconstruction, avoids the early convergence of the algorithm, and realizes a better balance between global search and local development, improves the calculation efficiency and the quality of the solution, and thus solves the reconstruction problem of the fuel cell ship power system in a more efficient and more accurate way.

[0021] The reconstruction is to change the current topology by reasonably changing the on-off state of the switch on the power grid, so as to remove overload, balance load, isolate faults and restore power.

[0022] Further, the position of the particle in the search space in the improved particle swarm algorithm corresponds to a set of solutions, that is, the on-off state of the switch.

[0023] In the application, the dimension of the particle is the number of switches, and the particle dimension is defined as 24, that is, each particle contains the on-off state of 24 switches, corresponding to the power supply condition of 24 loads in the power system. For important loads, the switch state can be 0 (off), 1 (normal path power supply) or 2 (backup path power supply), and for non-important loads, it can be 0 (off) or 1 (on).

[0024] Further, the improved particle swarm algorithm mainly consists of three parts: a logical mapping chaotic model for initializing candidate solutions, a Monte Carlo random equation for describing the position of the particle and introducing a diversity migration strategy, and a flow chart as shown in the figure. Figure 2 As shown in the figure, the specific steps of designing the improved particle swarm algorithm include: The logical mapping chaotic model is used to generate initial candidate solutions, which is beneficial to improve the uniformity of the candidate solutions, and the formula is as follows:

[0025] In the formula, is the candidate solution set, is the total number of candidate solutions, is the load serial number.

[0026] The improved particle swarm algorithm, quantum particle swarm algorithm in updating particle velocity and position, through the quantum probability to guide the search, can better avoid falling into local optimal solution, and has higher precision and robustness, the introduction of diversity migration strategy, further improve the global search ability of population, get more accurate search results. Using Monte Carlo random equation (Monte Carlo Method) to describe the particle position, using computer for statistical simulation and sampling, get the approximate optimal solution, can get the position equation of the particle in the potential well

[0027] Among them is a random number uniformly distributed on [0,1], is The characteristic length of the potential well. When the particle is in the multi-dimensional target search space, update the related variable equation as follows:

[0028]

[0029]

[0030]

[0031] In the formula is a random number uniformly distributed between [0,1], is the expansion-shrink coefficient, which decreases linearly from to , where , is the maximum number of iterations, is the current iteration number, is the average best position, is the th particle after the th iteration dimensional potential well length.

[0032] After transformation, the position equation of the particle is established as:

[0033] On the basis of quantum particle swarm algorithm, the specific implementation steps of diversity migration strategy are as follows: 1) Calculate the average Hamming distance in the population, select the particle with the smallest average Hamming distance as the migration object, and the average Hamming distance calculation formula of the particle is:

[0034] represents the total number of particles, and represents two particles.

[0035] According to the population representation, a particle with the minimum average Hamming value can be selected and added to the migration target list. In order to identify the particles to be migrated, the following two cases are listed: When there are no two or more particles with the same average Hamming in the population, the particle with the minimum average Hamming is selected as the migration candidate particle.

[0036] When there are two or more particles with the same average Hamming, the fitness values between the particle populations are compared, and the worst particle is selected as the migration object.

[0037] In order to remove the worst particle while increasing the diversity of the population, the particle with the maximum fitness value is replaced with the determined migration object.

[0038] Design simulation test model 1) According to the importance of the load, priority is given to three levels: 1st level load, i.e. critical load, requires uninterrupted power supply or unaffected power supply source when the ship is operating at sea, 2nd level load belongs to important load when the ship is operating, and 3rd level load, i.e. non-heavy load. For 1st and 2nd level loads, provide a backup power supply path to ensure the continuity of power supply, and 3rd level load is generally designed as a single path power supply.

[0039]

[0040] 2) For ease of analysis, the structure diagram shown in Figure 1 is converted into the system topology diagram shown in Figure 3 The fault simulation model is based on the fuel cell ship ring topology, in which FC1, FC2, FC3, and FC4 are four fuel cells with equal power levels, BAT1, BAT2, BAT3, and BAT4 are four lithium batteries with equal power levels, and L1-L24 are loads. The solid line in the figure is the normal power supply path, and the dashed line is the backup power supply path, which ensures the continuity of power supply for important loads in case of failure.

[0041] 3) Fault simulation and parameter setting, the data in Table 1 and the topology structure in Figure 3 are used to verify the algorithm, and it is assumed that branch 35 and branch 91 fail, as shown in Figures 4-6 , a fault reconstruction program is written by Matlab for simulation verification, and the parameter settings are shown in Table 2.

[0042]

[0043] 4) Algorithm integration and test: the improved particle swarm algorithm is integrated into the simulation model, the fault is set, and the optimal power supply path under the fault condition is found through the algorithm. The maximum recovery load, the minimum number of switch operations and the balanced load rate of each distribution board are considered as the objective function when the optimal solution is output.

[0044] Wherein, considering the priority of the load, the important load should be in working state, and the three-level load should be in working state as much as possible. Based on this, the first objective function is established as:

[0045] In the formula are the active power values of the first-level load, the second-level load and the third-level load respectively. are the number of different levels of loads respectively. represents the power supply state of the load, 0 represents the load unloading, and 1 represents the load loading. are the weights of different levels of loads, and the important load is required to be much larger than other loads. The weight expression is .

[0046] Considering the priority of the load, the number of switch operations should be as few as possible. Based on this condition, the second objective function is established as:

[0047] In the formula and or 1. and is the switch action state. The total number of important loads is represented, and the number of three-level loads is represented by .

[0048] The load rate of each regional power supply should be as balanced as possible, and based on this condition, the third objective function is established as:

[0049] In the formula, the number of power supplies is represented by , represents the actual power of the No. distribution board, represents the rated power of the No. distribution board.

[0050] Considering the above objective functions comprehensively, the final objective function is:

[0051] In the formula are the weight coefficients of each sub-objective respectively. It is required that . representing a maximum value of the objective function.

[0052] Based on the improved particle swarm algorithm in the fuel cell ship power system fault reconstruction simulation model, the switch code after 100 iterations of the algorithm is 211112011111111120111111, the number of switch operations is only 3 times, and the optimal solution tends to be stable after the second iteration, the value of fitness does not change, and the load loading rate of each distribution board stops changing after the second iteration. It can be seen that the method can quickly solve and compare the optimal solution in the reconstruction process, and has a faster convergence speed.

[0053] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can make equivalent replacements or changes within the technical range disclosed by the present application according to the technical solution and the inventive concept of the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for fault reconfiguration of a fuel cell ship power system, characterized in that... include: A simulation model of the ring power system of a fuel cell ship was built. The ship's load was classified according to its importance level, and backup power supply paths were configured for critical loads to control the continuity of power supply. The simulation model of the ring power system of the fuel cell ship simulated branch faults and power supply faults of the power system and tested various fault conditions. Design an improved particle swarm optimization algorithm: Initialize candidate solutions using a logistic mapping chaotic model, update candidate solutions using a quantum particle swarm optimization algorithm, and then introduce a diversity migration strategy to capture particles in different ranges in the population. Select migrating individuals based on fitness and population position information, where the individual with the smallest average Hamming distance in the population indicates the optimization direction of the iterative population. Replace particles that deviate from the population center range by comparing fitness values ​​and average Hamming distances between particles. An improved particle swarm optimization (PSO) algorithm is integrated into a simulation model of a fuel cell ship's ring power system. Each particle represents a solution to the fault reconfiguration problem, and the position of each particle in space corresponds to a specific set of switching states. Objective functions, namely, maximizing the restored load, minimizing the number of switching operations, and maximizing the power load rate of each region, are established. The three objective functions are weighted and summed to transform them into the final objective function, which is then optimized and solved. The improved PSO algorithm is used to reconfigure the power supply network under fault conditions, obtain the reconfiguration scheme, and complete the power system reconfiguration after the fault.

2. The method for fault reconfiguration of a fuel cell ship power system according to claim 1, characterized in that: The design of an improved particle swarm optimization algorithm includes: Initial candidate solutions are generated based on a logistic mapping chaotic model to improve the uniformity of candidate solutions: In the formula For the candidate solution set, The total number of candidate solutions. This is the load sequence number.

3. The method for fault reconfiguration of a fuel cell ship power system according to claim 1, characterized in that: The positions of particles in the improved particle swarm optimization algorithm are described using Monte Carlo random equations, and the position equations of particles in the potential well are obtained based on simulation and sampling methods: in It is a random number uniformly distributed on [0,1]. yes The characteristic length of the potential well, when the particle moves in a multidimensional target search space, updates the relevant variable equations, and the transformed particle position equation is established as follows: For the first The current optimal position of each particle. The contraction-expansion coefficient, This represents the average optimal position.

4. The method for fault reconfiguration of a fuel cell ship power system according to claim 1, characterized in that: When introducing a diversity migration strategy: calculate the average Hamming distance in the population, and select the particle with the smallest average Hamming distance as the migration target. The formula for calculating the average Hamming distance of a particle is: Represents the total number of particles. and This represents two particles.

5. Select the particle with the lowest mean Hamming value based on the population representation and add it to the migration target list. When there are no two or more particles with the same mean Hamming in the population, select the particle with the lowest mean Hamming as the migration candidate particle. When there are two or more particles with the same mean Hamming, compare the fitness values ​​between the particle populations and select the worst particle as the migration target.

6. A method for fault reconfiguration of a fuel cell ship power system according to claim 4, characterized in that: While increasing population diversity, the worst particles are removed, and the determined migration targets are replaced by the particles with the highest fitness values.

7. The method for fault reconfiguration of a fuel cell ship power system according to claim 1, characterized in that: Based on load priority, critical loads should be kept active, and tertiary loads should be kept active as much as possible. Therefore, the first objective function is established as follows: In the formula These are the active power values ​​for primary, secondary, and tertiary loads, respectively. These represent the number of load levels at different rates. This represents the power supply status of the load; 0 indicates the load is unloaded, and 1 indicates the load is loaded. These are the weights assigned to different load levels, with critical loads required to have significantly higher weights than other loads. The weighting expression is as follows: Based on load priority, if the number of switching operations is kept to a minimum, then the second objective function is established as follows: In the formula and Or 1, and This indicates the switch operation status and the total number of critical loads. This indicates that the number of three loads is determined by... express; To balance the load factor of the power supply in each area, a third objective function is established based on this condition: The number of power sources in the formula is... express, express The actual power of the No. 1 distribution board express Rated power of the No. 1 distribution board; Construct the final objective function based on the first, second, and third objective functions: In the formula These are the weight coefficients for each sub-objective, and the requirements are... , This represents the maximum value of the objective function.

Images